Two upcoming talks of interest Shared Circuits and the understanding of actions, sensations, and emotions. Christian Keysers 12:30 – 2 5 A ICSI, 4/7/2006 Brain Mechanisms of meaning access: From Action to Abstr-Action F. Pulvermuller, Cambridge 11-12:30, 6 th floor, ICSI
45
Embed
Two upcoming talks of interest Shared Circuits and the understanding of actions, sensations, and emotions. Christian Keysers 12:30 – 2 5 A ICSI, 4/7/2006.
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Two upcoming talks of interest
Shared Circuits and the understanding of actions, sensations, and emotions.
Christian Keysers 12:30 – 2 5 A ICSI, 4/7/2006
Brain Mechanisms of meaning access: From Action to Abstr-Action
F. Pulvermuller, Cambridge 11-12:30, 6th floor, ICSI
Lecture Overview
Metaphors Primary Metaphors Complex Metaphors
A computational Model of Event Structure Applying the Model to understanding newspaper
articles. Demo
Extensions and Scalable Inference
Task Interpret simple discourse fragments/blurbs
France fell into recession. Pulled out by Germany Economy moving at the pace of a Clinton jog. US Economy on the verge of falling back into recession after
moving forward on an anemic recovery. Indian Government stumbling in implementing Liberalization
plan. Moving forward on all fronts, we are going to be ongoing and
relentless as we tighten the net of justice. The Government is taking bold new steps. We are loosening
the stranglehold on business, slashing tariffs and removing obstacles to international trade.
Basic Components
An fine-grained executing model of action and events (X-schemas).
A simulation of connected embodied x-schemas using a controller x-schema
A representation of the domain/frames (DBN’s) that supports spreading activation
A model of metaphor maps that project bindings from source to target domains.
An Active Model of Events
Computationally, actions and events are coded in active representations called x-schemas which are extensions to Stochastic Petri nets.
x-schemas are fine-grained action and event representations that can be used for monitoring and control as well as for inference.
The controller schema provides a compositional mechanism to compose events through activation, inhibition, and modification
Simulation hypothesis
We understand utterances by mentally simulating their content.
– Simulation exploits some of the same neural structures activated during performance, perception, imagining, memory…
– Linguistic structure parameterizes the simulation.• Language gives us enough information to simulate
Inference from Domain knowledge
Language understanding never occurs in a vacuum – in making sense of an utterance we use both our general
experience of the world and our beliefs about the current situation.
X-schemas describe our embodied knowledge of action and of processes are used in comprehending language.
The programs that interpret news stories must also make inferences from descriptive (Frame and Domain) knowledge.
The Target Domain Simple knowledge about Economics
Factual (US is a market economy) Correlational (High Growth => High Inflation)
Key Requirement: Must combine background knowledge of economics with
inherent structure and constraints of the target domain with inferential products of metaphoric (and other) projections from multiple source domains.
Must be able to compute the global impact of new observations (from direct input as well as metaphoric inferences). Such inference should model spreading activation (parallel, top-down and bottom up)
Modeling Spreading Activation
Traditional theories of meaning have focused entirely on logical deduction as a model of understanding. Although much has been learned from this approach, it only covers a
small fraction of the kinds of inferences that people draw when understanding language.
From our neural perspective, inference is better seen as a process of quantitatively combining evidence in context to derive the most likely conclusions.
When you hear or read something new, your brain’s spreading activation mechanisms automatically connect it to related information. Strength of connection Strength of activation
A computational model: Bayes Nets
At the computational level, Bayes Networks capture the best fit character of neural inference allows us to model a much wider range of language behavior.
BN are a computational formalism that is the best available approximation to neural spreading activation.
In this lecture, we will combine bayes networks active schemas
in a computational model of how people understand the meaning of news stories about economics.
Bayes Networks
Expoits conditional independence requiring only local conditional beliefs.Basic operation is conditioning in the presence of evidence. Supports Multiple inference types
Forward Inter-causalBackward
Example: AlarmExample: Alarm
Five state featuresFive state features A: Alarm A: Alarm B: BurglaryB: Burglary E: EarthquakeE: Earthquake J: JohnCallsJ: JohnCalls M: MaryCallsM: MaryCalls
A Simple Bayes NetA Simple Bayes Net
Burglary Earthquake
Alarm
MaryCallsJohnCalls
causes
effects
Directed acyclicgraph (DAG)
Intuitive meaning of arrowfrom x to y: “x has direct influence on y”
Nodes are feature-value structs
Assigning Probabilities to Assigning Probabilities to RootsRoots
Burglary Earthquake
Alarm
MaryCallsJohnCalls
P(B)P(B)
0.000.0011
P(E)P(E)
0.000.0022
Conditional Probability TablesConditional Probability Tables
BB EE P(A|P(A|……))
TTTTFFFF
TTFFTTFF
0.950.950.940.940.290.290.0010.001
Burglary Earthquake
Alarm
MaryCallsJohnCalls
P(B)P(B)
0.000.0011
P(E)P(E)
0.000.0022
Size of the CPT for a node with k parents: 2k
Conditional Probability Conditional Probability TablesTables
BB EE P(A|P(A|……))
TTTTFFFF
TTFFTTFF
0.950.950.940.940.290.290.0010.001
Burglary Earthquake
Alarm
MaryCallsJohnCalls
P(B)P(B)
0.000.0011
P(E)P(E)
0.000.0022
AA P(J|…)P(J|…)
TTFF
0.900.900.050.05
AA P(M|P(M|…)…)
TTFF
0.700.700.010.01
What the BN MeansWhat the BN Means
BB EE P(A|P(A|……))
TTTTFFFF
TTFFTTFF
0.950.950.940.940.290.290.0010.001
Burglary Earthquake
Alarm
MaryCallsJohnCalls
P(B)P(B)
0.000.0011
P(E)P(E)
0.000.0022
AA P(J|P(J|…)…)
TTFF
0.900.900.050.05
AA P(M|P(M|…)…)
TTFF
0.700.700.010.01
P(x1,x2,…,xn) = i=1,…,nP(xi|
Parents(Xi))
Calculation of Joint Calculation of Joint ProbabilityProbability
BB EE P(A|P(A|……))
TTTTFFFF
TTFFTTFF
0.950.950.940.940.290.290.0010.001
Burglary Earthquake
Alarm
MaryCallsJohnCalls
P(B)P(B)
0.000.0011
P(E)P(E)
0.000.0022
AA P(J|…)P(J|…)
TTFF
0.900.900.050.05
AA P(M|P(M|…)…)
TTFF
0.700.700.010.01
P(JMABE)= P(J|A)P(M|A)P(A|B,E)P(B)P(E)= 0.9 x 0.7 x 0.001 x 0.999 x 0.998= 0.00062
What the BN EncodesWhat the BN Encodes
Each of the beliefs Each of the beliefs JohnCalls and JohnCalls and MaryCalls is MaryCalls is independent of independent of Burglary and Burglary and Earthquake given Earthquake given Alarm or Alarm or AlarmAlarm
The beliefs JohnCalls The beliefs JohnCalls and MaryCalls are and MaryCalls are independent given independent given Alarm or Alarm or AlarmAlarm
Burglary Earthquake
Alarm
MaryCallsJohnCalls
For example, John doesnot observe any burglariesdirectly
What the BN EncodesWhat the BN Encodes
Each of the beliefs Each of the beliefs JohnCalls and JohnCalls and MaryCalls is MaryCalls is independent of independent of Burglary and Burglary and Earthquake given Earthquake given Alarm or Alarm or AlarmAlarm
The beliefs JohnCalls The beliefs JohnCalls and MaryCalls are and MaryCalls are independent given independent given Alarm or Alarm or AlarmAlarm
Burglary Earthquake
Alarm
MaryCallsJohnCalls
For instance, the reasons why John and Mary may not call if there is an alarm are unrelated
Note that these reasons couldbe other beliefs in the network.The probabilities summarize thesenon-explicit beliefs
D-SeparationD-Separation
Say we want to know the probability of Say we want to know the probability of some variable (e.g. JohnCalls) given some variable (e.g. JohnCalls) given evidence on another (e.g. Alarm). What evidence on another (e.g. Alarm). What variables are relevant to this calculation?variables are relevant to this calculation?
I.e.: Given an arbitrary graph G = (V,E), I.e.: Given an arbitrary graph G = (V,E), is Xis XA A independent of Xindependent of XBB|X|XCC for some A,B, for some A,B, and C?and C?
The answer can be read directly off the The answer can be read directly off the graph, using a notion called graph, using a notion called D-D-separationseparation
What can Bayes nets be used for? Posterior probabilities
Probability of any event given any evidence
Most likely explanationScenario that explains evidence
Rational decision makingMaximize expected utilityValue of Information
Markov Random FieldBoltzmann machineIsing modelMax-ent modelLog-linear models
(Bayesian belief nets) (Markov nets)
Bayes Nets and Human Probabilistic Inference Our use of Bayes Networks will be to model how people
reason about uncertain events, such as those in economics and politics.
We know that people do reason probabilistically, but also that they do not always act in accord with the formal laws of probability. Daniel Kahneman won the 2002 Nobel Prize largely for his
work with Amos Tversky explaining many of the limitations of human probabilistic reasoning. Some of the limitations are obvious, e.g. the calculations might be just too complex.
But some are much deeper involving the way a question is stated, a preference for avoiding loss, and some basic misperceptions about large and small probabilities.
Bayes nets only approximate the underlying evidential neural computation, but they are by far the best available model.
Economic State [recession,nogrowth,lowgrowth,higrowth]
Goal
Policy
Outcome
Difficulty
A Simple DBN for the target domain
[Liberalization, Protectionism]
[Free Trade, Protection ]
[Success, failure]
[present, absent]
T0 T1
Probabilistic inference
Filtering P(X_t | o_1…t,X_1…t) Update the state based on the observation sequence and state
set MAP Estimation
Argmaxh1…hnP(X_t | o_1…t, X_1…t) Return the best assignment of values to the hypothesis
variables given the observation and states Smoothing
P(X_t-k | o_1…t, X_1…t) modify assumptions about previous states, given observation
sequence and state set Projection/Prediction/Reachability
P(X_t+k | o_1..t, X_1..t)
Metaphor Maps
Static Structures that project bindings from source domain f- struct to target domain Bayes net nodes by setting evidence on the target network.
Different types of maps PMAPS project X- schema Parameters to abstract domains OMAPS connect roles between source and target domain SMAPS connect schemas from source to target domains.
ASPECT is an invariant in projection.
FRAME Ec_PolicySUBCASE OF ActionROLES
Degree of Progress
FRAME JourneySUBCASE OF Self MotionROLES
Rate of Motion
MAP ProgressISRate
map-type <- METAPHOR
tgt srcPAIRS
Lecture Overview
MetaphorsPrimary MetaphorsComplex Metaphors
A computational Model of Event Structure Applying the Model to understanding
newspaper articles.Demo
Extensions and Scalable Inference
I/O as Feature Structures Indian Government stumbling in implementing
liberalization plan
KARMA DEMO
• SOURCE DOMAINS: MOTION, HEALTH
• TARGET DOMAINS: INTERNATONAL ECONOMICS
• METAPHOR MAPS: EVENT STRUCTURE METAPHOR
Results Model was implemented and tested on discourse fragments
from a database of 50 newspaper stories in international economics from standard sources such as WSJ, NYT, and the Economist. Results show that motion terms are often the most effective method to provide the following types of information about abstract plans and actions. Information about uncertain events and dynamic changes in
goals and resources. (sluggish, fall, off-track, no steam) Information about evaluations of policies and economic actors and
communicative intent (strangle-hold, bleed). Communicating complex, context-sensitive and dynamic